Modulation Recognition Research For Commonly Used Digital Signal

Modulation recognition of Digital Communication Signal has always been a research focus for countries, which has a wide range of applications in military and civilian communication fields. In recent years, the researchers combined with intelligent pattern recognition methods such as neural networks, support vector machines, kernel methods, fuzzy, etc. proposed a lot of modulation recognition methods. By selecting and extracting some original modulation characteristics, and constructing a new applicable feature, this paper put forward a new Statistics Pattern Recognition Algorithm which based on minimum distance and fuzzy matching.The algorithm does not just use a single threshold of parameters to simply classify signals, but rather a multi-parameter joint identification. To identify the inter-class, using a minimum distance classification algorithm, can effectively distinguish ASK, PSK, FSK, QAM signals when the SNR is greater than 8db. To identify the type, using fuzzy matching algorithm and combined with the threshold to identify QAM signal, can effectively distinguish the decimal of ASK, PSK, FSK, QAM signals when the SNR is greater than 8db. To solve the problem that FSK, PSK signal is difficult to identify in low SNR conditions, using second moment, fourth moment algorithm to estimate the SNR of noise contaminated FSK, PSK signal, and studied the SNR impact to Parameters so as to provides the theoretical basis to modify the characteristics of the system in real time. A large number of computer simulation experiments show that ASK,PSK,FSK and QAM signals recognition rate of correct classification are close to 98% when the SNR is greater than 8db. Compared with the existing algorithms, the algorithm's classification performance has further improved.